Topic Notes
Visual Perception
CS 7450 - Information Visualization
August 29, 2012
John Stasko
Agenda
• Visual perception
 Pre-attentive processing
 Color
 Etc.
Fall 2012
CS 7450
2
1
Semiotics
• The study of symbols and how they
convey meaning
• Classic book:
 J. Bertin, 1983, The Semiology of Graphics
Fall 2012
CS 7450
3
Related Disciplines
• Psychophysics
 Applying methods of physics to measuring
human perceptual systems
How fast must light flicker until we perceive it as
constant?
What change in brightness can we perceive?
• Cognitive psychology
 Understanding how people think, here, how it
relates to perception
Fall 2012
CS 7450
4
2
Perceptual Processing
• Seek to better understand visual
perception and visual information
processing
 Multiple theories or models exist
 Need to understand physiology and cognitive
psychology
Fall 2012
CS 7450
5
One (simple) Model
• Two stage process
 Parallel extraction of low-level properties of
scene
 Sequential goal-directed processing
Stage 1
Ware 2000
Fall 2012
Stage 2
Early, parallel
detection of
color, texture,
shape, spatial
attributes
CS 7450
Serial processing of
object identification (using
memory) and spatial layout,
action
6
3
Stage 1 - Low-level, Parallel
• Neurons in eye & brain responsible for different
kinds of information
 Orientation, color, texture, movement, etc.
•
•
•
•
Arrays of neurons work in parallel
Occurs “automatically”
Rapid
Information is transitory, briefly held in iconic
store
• Bottom-up data-driven model of processing
• Often called “pre-attentive” processing
Fall 2012
CS 7450
7
Stage 2 - Sequential, Goal-Directed
• Splits into subsystems for object recognition and for
interacting with environment
• Increasing evidence supports independence of systems
for symbolic object manipulation and for locomotion &
action
• First subsystem then interfaces to verbal linguistic
portion of brain, second interfaces to motor systems
that control muscle movements
Fall 2012
CS 7450
8
4
Stage 2 Attributes
• Slow serial processing
• Involves working and long-term memory
• More emphasis on arbitrary aspects of
symbols
• Top-down processing
Fall 2012
CS 7450
9
Preattentive Processing
• How does human visual system analyze
images?
 Some things seem to be done preattentively,
without the need for focused attention
 Generally less than 200-250 msecs (eye
movements take 200 msecs)
 Seems to be done in parallel by low-level
vision system
Drawn from
C. Healey web article
Fall 2012
CS 7450
10
5
How Many 3’s?
1281768756138976546984506985604982826762
9809858458224509856458945098450980943585
9091030209905959595772564675050678904567
8845789809821677654876364908560912949686
Fall 2012
CS 7450
11
How Many 3’s?
1281768756138976546984506985604982826762
9809858458224509856458945098450980943585
9091030209905959595772564675050678904567
8845789809821677654876364908560912949686
Fall 2012
CS 7450
12
6
What Kinds of Tasks?
• Target detection
 Is something there?
• Boundary detection
 Can the elements be grouped?
• Counting
 How many elements of a certain type are
present?
Fall 2012
CS 7450
13
Example
• Determine if a red circle is present
• (2 sides of the room)
Fall 2012
CS 7450
14
7
Hue
Can be done rapidly (preattentively) by people
Surrounding objects called “distractors”
Fall 2012
CS 7450
15
Example
• Determine if a red circle is present
Fall 2012
CS 7450
16
8
Shape
Can be done preattentively by people
Fall 2012
CS 7450
17
Example
• Determine if a red circle is present
Fall 2012
CS 7450
18
9
Hue and Shape
• Cannot be done preattentively
• Must perform a sequential search
• Conjuction of features (shape and hue) causes it
Fall 2012
CS 7450
19
Example
• Is there a boundary in the display?
Fall 2012
CS 7450
20
10
Fill and Shape
• Left can be done preattentively since each group
contains one unique feature
• Right cannot (there is a boundary!) since the two
features are mixed (fill and shape)
Fall 2012
CS 7450
21
Example
• Is there a boundary in the display?
Fall 2012
CS 7450
22
11
Hue versus Shape
Left: Boundary detected preattentively based
on hue regardless of shape
Right: Cannot do mixed color shapes preattentively
Fall 2012
CS 7450
23
Example
• Is there a boundary?
Fall 2012
CS 7450
24
12
Hue versus brightness
Left: Varying brightness seems to interfere
Right: Boundary based on brightness can be
done preattentively
Fall 2012
CS 7450
25
Example Applet
• Nice on-line tutorial and example applet
 http://www.csc.ncsu.edu/faculty/healey/PP/index.html
 Chris Healey, NC State
 Prior pictures taken from site
Fall 2012
CS 7450
26
13
Preattentive Features
• Certain visual forms lend themselves to
preattentive processing
• Variety of forms seem to work
Fall 2012
CS 7450
27
Textons
1. Elongated blobs
2. Terminators
3. Crossings of lines
Fall 2012
All detected early
CS 7450
28
14
3-D Figures
3-D visual reality has an influence
Fall 2012
CS 7450
29
Emergent Features
Fall 2012
CS 7450
30
15
Potential PA Features
length
width
size
curvature
number
terminators
intersection
closure
Fall 2012
hue
intensity
flicker
direction of motion
binocular lustre
stereoscopic depth
3-D depth cues
lighting direction
CS 7450
31
Discussion
• What role does/should preattentive
processing play in information
visualization?
Fall 2012
CS 7450
32
16
Gestalt Laws
• Background
 German psychologists, early 1900’s
 Attempt to understand pattern perception
 Founded Gestalt school of psychology
 Provided clear descriptions of many basic
perceptual phenomena
 Gestalt Laws of Pattern Perception
Fall 2012
CS 7450
33
Gestalt Laws
• Proximity
 Things close together are perceptually grouped
together
• Similarity
 Similar elements get grouped together
• Connectedness
 Connecting different objects by lines unifies them
• Continuity
 More likely to construct visual entities out of smooth,
continuous visual elements
Fall 2012
CS 7450
34
17
Gestalt Laws
• Symmetry
 Symmetrical patterns are perceived more as a whole
• Closure
 A closed contour is seen as an object
• Relative Size
 Smaller components of a pattern as perceived as
objects
• Figure & Ground
 Figure is foreground, ground is behind
Fall 2012
CS 7450
35
Key Perceptual Properties
•
•
•
•
Brightness
Color
Texture
Shape
Fall 2012
CS 7450
36
18
Luminance/Brightness
• Luminance
 Measured amount of light coming from some
place
• Brightness
 Perceived amount of light coming from source
Fall 2012
CS 7450
37
Brightness
• Perceived brightness is non-linear function
of amount of light emitted by source
 Typically a power function
 S = aIn
S - sensation
I - intensity
• Very different on screen versus paper
Fall 2012
CS 7450
38
19
Grayscale
• Probably not best way to encode data
because of contrast issues
 Surface orientation and surroundings matter
a great deal
 Luminance channel of visual system is so
fundamental to so much of perception
We can get by without color discrimination, but not
luminance
Fall 2012
CS 7450
39
Color
• Sensory response to electromagnetic
radiation in the spectrum between
wavelengths 0.4 - 0.7 micrometers
10-6
gamma
Fall 2012
10-1
0.5
105
108
ultraviolet
visible
microwave
tv
CS 7450
40
20
Color Models
• HVS model
 Hue - what people think of color
 Value - light/dark, ranges black<-->white
 Saturation - intensity, ranges hue<-->gray
white
Value
Hue
Saturation
Fall 2012
black
CS 7450
41
How Not to Use Color
http://www.thedailyshow.com/video/index.jhtml?videoId=156230&title=full-color-coverage
Fall 2012
CS 7450
42
21
Color Categories
• Are there certain
canonical colors?
 Post & Greene ‘86
had people name
different colors on a
monitor
 Pictured are ones
with > 75%
commonality
From Ware ‘04
Fall 2012
CS 7450
43
Using Mechanical Turk
http://blog.doloreslabs.com/2008/03/where-does-blue-end-and-red-begin/
Fall 2012
CS 7450
44
22
Maybe Not All the Same?
http://www.eversostrange.com/2011/08/11/himba-colours-differently/
Fall 2012
CS 7450
45
Luminance
• Important for fg-bg colors to differ in
brightness
Hello, here is some text. Can you read what it says?
Hello, here
Hello, here
Hello, here
Hello, here
Hello, here
is
is
is
is
is
some
some
some
some
some
text.
text.
text.
text.
text.
Can
Can
Can
Can
Can
you read what
you read what
you read what
you read what
you read what
it says?
it says?
it says?
it says?
it says?
Hello, here is some text. Can you read what it says?
Fall 2012
CS 7450
46
23
Color for Categories
• Can different colors be used for
categorical variables?
 Yes (with care)
 Ware’s suggestion: 12 colors
red, green, yellow, blue, black, white, pink, cyan,
gray, orange, brown, purple
From Ware ‘04
Fall 2012
CS 7450
47
Color for Sequences
Can you order these (low->hi)
Fall 2012
CS 7450
48
24
Possible Color Sequences
Gray scale
Full spectral scale
Fall 2012
Single sequence
part spectral scale
Single sequence
single hue scale
Double-ended
multiple hue scale
CS 7450
49
HeatMap
•http://screening.nasdaq.com/heatmaps/heatmap_100.asp
Fall 2012
CS 7450
50
25
ColorBrewer
Help with selecting
colors for maps
http://colorbrewer2.org/
Fall 2012
CS 7450
51
Color Purposes
• Call attention to specific data
• Increase appeal, memorability
• Increase number of dimensions for
encoding data
 Example, Ware and Beatty ‘88
x,y - variables 1 & 2
amount of r,g,b - variables 3, 4, & 5
Fall 2012
CS 7450
52
26
Using Color
• Modesty! Less is more
• Use blue in large regions, not thin lines
• Use red and green in the center of the
field of view (edges of retina not sensitive
to these)
• Use black, white, yellow in periphery
• Use adjacent colors that vary in hue &
value
Fall 2012
CS 7450
53
Using Color
• For large regions, don’t use highly saturated
colors (pastels a good choice)
• Do not use adjacent colors that vary in amount
of blue
• Don’t use high saturation, spectrally extreme
colors together (causes after images)
• Use color for grouping and search
• Beware effects from adjacent color regions (my
old house - example)
Fall 2012
CS 7450
54
27
Article Discussion
http://www.b-eye-network.com/newsletters/ben/2235
Fall 2012
CS 7450
55
Good Color Advice
Maureen Stone’s website
Many references and links
She frequently offers
tutorials about color at
conferences
http://www.stonesc.com
Fall 2012
CS 7450
56
28
Color Challenge
Fall 2012
http://color.method.ac/
CS 7450
57
Texture
• Appears to be combination of
 orientation
 scale
 contrast
• Complex attribute to analyze
Fall 2012
CS 7450
58
29
Shape, Symbol
• Can you develop a set of unique symbols
that can be placed on a display and be
rapidly perceived and differentiated?
• Application for maps, military, etc.
• Want to look at different preattentive
aspects
Fall 2012
CS 7450
59
Glyph Construction
• Suppose that we use two different visual
properties to encode two different variables in a
discrete data set
 color, size, shape, lightness
• Will the two different properties interact so that
they are more/less difficult to untangle?
 Integral - two properties are viewed holistically
 Separable - Judge each dimension independently
Fall 2012
CS 7450
60
30
Integral-Separable
• Not one or other, but along an axis
Integral
Separable
red-green
red-green
shape height
shape
color
direction motion
color
color
x,y position
yellow-blue
black-white
shape width
size
size
shape
shape
direction motion
size, shape, color
Ware ‘04
Fall 2012
CS 7450
61
Encodings
• When you want to communicate one type
of variable, which visual property should
you use?
Fall 2012
CS 7450
62
31
Accuracy ranking of
quantitative perceptual
tasks
Ranking of perceptual
tasks
From Jock Mackinlay’s 1986 thesis
Fall 2012
CS 7450
63
Change Blindness
• Is the viewer able to perceive changes
between two scenes?
 If so, may be distracting
 Can do things to minimize noticing changes
• Fun examples
 Static pictures (Ron Rensink, UBC)
http://www.psych.ubc.ca/~rensink/flicker/download/
 Videos (Dan Simons, Illinois)
http://viscog.beckman.uiuc.edu/djs_lab/demos.html
Fall 2012
CS 7450
64
32
Optical Illusions
Fall 2012
CS 7450
65
Stage 2
• Missing here!
• Object recognition and locomotion/action
• Maybe in the future… :^)
Fall 2012
CS 7450
66
33
Great Book
Information Visualization
Perception for Design
2nd edition
Colin Ware
Morgan Kaufmann
Fall 2012
CS 7450
Design Project
67
More details on website
• Group of 2-4 students
• Understand problem, design, build
• You pick the topic/domain/data
 Absolutely crucial!!!
 NY Times vizs are nice examples
 Be creative!
• First milestone: Teams and topics in 2.5
weeks (Sep 17th)
Fall 2012
CS 7450
68
34
HW 2
• Due Wednesday
• Questions?
Fall 2012
CS 7450
69
Upcoming
• Labor Day holiday
 No class
• Cognitive Issues
 Papers to read
Norman book chapter
Liu et al ‘08
Fall 2012
CS 7450
70
35
Sources Used
Healey website and article
http://www.csc.ncsu.edu/faculty/healey/PP/index.html
Marti Hearst SIMS 247 lectures
C. Ware, Information Visualization
Fall 2012
CS 7450
71
36